ScSR
所属分类：图形图像处理
开发工具：matlab
文件大小：26991KB
下载次数：480
上传日期：2013-04-02 09:09:31
上 传 者：
dfsqjzz
说明： Jianchao Yang 的基于稀疏表示的单幅图像重建的原始代码，先将高低训练图像分块，再将块训练成高低字典，将测试图像映射到低字典上，得到系数，再乘以高子典就得到最后的图像。对学习超分辨率同学的参考作用很大。
(This is the original matlab code for super resolution by Jianchao Yang 。The method is sparse represent based on the overcomplete dictionary。)
文件列表:
ScSR (0, 2013-04-02)
ScSR\Data (0, 2011-03-07)
ScSR\Data\Testing (0, 2013-04-02)
ScSR\Data\Testing\gnd.bmp (196662, 2011-01-10)
ScSR\Data\Testing\input.bmp (49206, 2011-01-18)
ScSR\Data\Testing\result.bmp (196662, 2011-01-28)
ScSR\Data\Training (0, 2013-04-02)
ScSR\Data\Training\t1.bmp (104246, 2007-10-14)
ScSR\Data\Training\t11.bmp (118614, 2007-10-14)
ScSR\Data\Training\t12.bmp (40990, 2007-10-14)
ScSR\Data\Training\t13.bmp (101010, 2007-10-14)
ScSR\Data\Training\t14.bmp (76990, 2007-10-14)
ScSR\Data\Training\t16.bmp (83894, 2007-10-14)
ScSR\Data\Training\t17.bmp (69786, 2007-10-14)
ScSR\Data\Training\t18.bmp (55782, 2007-10-14)
ScSR\Data\Training\t19.bmp (96174, 2007-10-14)
ScSR\Data\Training\t2.bmp (92418, 2007-10-14)
ScSR\Data\Training\t20.bmp (18462, 2007-10-14)
ScSR\Data\Training\t21.bmp (41346, 2007-10-14)
ScSR\Data\Training\t22.bmp (59718, 2007-10-14)
ScSR\Data\Training\t23.bmp (46902, 2007-10-14)
ScSR\Data\Training\t24.bmp (37290, 2007-10-14)
ScSR\Data\Training\t25.bmp (125190, 2007-10-14)
ScSR\Data\Training\t26.bmp (58742, 2007-10-14)
ScSR\Data\Training\t27.bmp (126030, 2007-10-14)
ScSR\Data\Training\t28.bmp (92550, 2007-10-14)
ScSR\Data\Training\t3.bmp (89334, 2007-10-14)
ScSR\Data\Training\t30.bmp (61662, 2007-10-14)
ScSR\Data\Training\t31.bmp (101670, 2007-11-09)
ScSR\Data\Training\t32.bmp (86646, 2007-11-09)
ScSR\Data\Training\t34.bmp (66362, 2007-11-09)
ScSR\Data\Training\t35.bmp (125306, 2007-11-09)
ScSR\Data\Training\t36.bmp (114102, 2007-11-23)
ScSR\Data\Training\t37.bmp (312390, 2007-11-22)
ScSR\Data\Training\t38.bmp (200054, 2007-11-22)
ScSR\Data\Training\t39.bmp (197350, 2007-11-22)
ScSR\Data\Training\t4.bmp (128646, 2007-10-14)
ScSR\Data\Training\t40.bmp (200082, 2007-11-22)
ScSR\Data\Training\t42.bmp (202814, 2007-11-23)
ScSR\Data\Training\t43.bmp (157134, 2007-11-23)
... ...
*****************************************************************
* Demo Codes For Image Super-resolution via Sparse *Representation
*****************************************************************
Reference
J. Yang et al. Image super-resolution as sparse representation of raw image patches. CVPR 2008.
J. Yang et al. Image super-resolution via sparse representation. IEEE Transactions on Image Processing, Vol 19, Issue 11, pp2861-2873, 2010
For any problems, send email to jyang29@uiuc.edu
================================================================
Demo_SR.m: demo code for image super-resolution via sparse recovery
1. The demo code is for upscaling factor of 2. For larger magnification factors, run the function "ScSR.m" multiple times. Note the code is a little different from what presented in the TIP10 paper. Please find the previous codes and results in folder "Previous".
2. Two pre-trained dictionaries are provided in directory "Dictionary". The dictionaries are for zoom factor of 2. You can train your own dictionary based on function "Demo_Dictionary_Training.m" talked below.
================================================================
Demo_Dictionary_Training.m: demo code for training the dictionary
1. If you want to train your own dictionary, replace the training images in subfolder "Data/Training" by yours.
2. You need to inspect the statistics of your sampled patches to prune those smooth patches.
近期下载者：
相关文件：
收藏者：